In recent years, researches for deducing past variation or presuming future change of the global environment by the determination of hydrogen and oxygen stable isotopes of water such as rainwater, soil water, or polar ice core are actively conducted, and isotope ratio mass spectrometers for implementing this are also under development (See, for example, Patent document 1). For example, meteoric water in the subtropical or temperate regions show characteristic seasonal variations in their isotopic composition, which can provide important information as suggesting that the source of the meteoric water and/or the temperature at which water vapor are condensed change seasonally. This also provides an index for tracking changes in flow condition of groundwater when an assessment of construction project is conducted in association with huge undertaking.
Stable isotopic composition of water is usually analyzed in a separate manner such that a hydrogen isotope is determined as a form of hydrogen gas and an oxygen isotope is determined principally as a form of carbon dioxide gas or exceptionally as a form of oxygen gas because water vapor is an adsorptive gas having high affinity to metals, which will reduce the precision of accuracy for the determination significantly. To determine hydrogen stable isotopic composition of water using isotope ratio mass spectrometer (hereafter IRMS), a reduction method, an equilibrium method, and the like have been used. The reduction method is performed by reacting water molecules with metals such as uranium or zinc at the high temperature in vacuum to generate hydrogen gases. These hydrogen gases introduce directly to the IRMS (See, for example, non-patent document 1). On the other hand, the equilibrium method is performed by equilibrating water with arbitrary hydrogen gas at a constant temperature under the presence of platinum catalyst. This reaction is represented in Chemical formula 1

The reduction method described in the non-patent document 1, however, entailed the problems that a vacuum device is required, the platinum catalyst used for the reaction is expensive, and it is difficult to prevent the metal surface from getting oxidized. Also the equilibrium method described in the non-patent document 1 had a problem that it requires 1 hour or longer time to reach the equilibration for hydrogen isotope between water and hydrogen gas as represented in Chemical formula 1, and the automated equipment for this reaction is expensive.
On the other hand, to determine oxygen stable isotopic composition of water, an equilibrium method and an oxidation method were performed conventionally. The equilibrium method is performed by equilibrating water with arbitrary carbon dioxide gas at a constant temperature to obtain the oxygen isotope equilibrium between them, as represented by Chemical formula 2 (See, for example, non-patent document 2). The oxidation method is performed by reacting water with fluorides such as bromine pentafluoride to extract oxygen gas (See, for example, non-patent document 2).2H218O+C16O22H216O+C18O16O  [Chemical formula 2]
The conventional equilibrium method described in the non-patent document 2, had the problems that it requires 10 hour or longer time to reach the isotopic equilibration for oxygen isotope between water and carbon dioxide as represented in Chemical formula 2. It is impossible to analyze 17O using a conventional IRMS which has relatively low resolution for separating the small difference of masses of isotopic isomer between 12C17O16O and 13C16O16O. The automated equipment for this reaction is expensive. The conventional oxidation method described in the non-patent document 2 had the problems that a vacuum device is required and handling of bromine pentafluoride and the like is difficult and dangerous.
There is an alternative method to determine hydrogen and oxygen isotopic composition, which is a specific method using the tunable diode-laser spectroscopy. Since it is free from the constraint that those having equivalent masses cannot be determined as is the case of the mass spectrometry, it is possible to determine 17O. However, laser spectroscopy for the determination of stale isotopic composition of water is a very specialized analytical method and is still inferior in accuracy to conventional isotope ratio mass spectrometry at this time.
For addressing these problems, another method attempts to determine the contents 17O by extracting oxygen gas by electrolysis in a single electrolytic chamber which contains platinum serving as an anode and an oxygen-free carbon electrode serving as a cathode in an copper sulfate solution (See, for example, non-patent document 3).
According to the method described in the non-patent document 3, however, it is necessary to add an electrolyte, and addition of electrolyte may cause the analytical error in isotope ratio mass spectrometry or cause the failure of the IRMS. Furthermore, in the above method, it is possible to extract only oxygen but not a hydrogen isotope. Additionally, it takes 40 minutes to extract enough amount of oxygen gas for use in mass spectrometry.
In addition, conventionally, since oxygen gas and hydrogen gas can be extracted only as gas mixture from same sample water, it is necessary to provide an analytical procedure to separate oxygen gas and hydrogen gas from obtained mixture of oxygen gas and hydrogen gas in order to determine oxygen stable isotopic composition and/or hydrogen stable isotopic composition.
The following are prior art references related to the present invention.
Patent document 1: Japanese Patent Gazette No. 3048146
Non-patent document 1: New Experimental Chemistry Series 10, Space and Geo Chemistry, 1976, MARUZEN, p. 485.
Non-patent document 2: New Experimental Chemistry Series 10, Space and Geo Chemistry, 1976, MARUZEN, p. 486-p. 491.
Non-patent document 3: H. A J. MEIJER and W. J. LI, THE USE OF ELECTROLYSIS FOR ACCURATE δ17O AND δ18O ISOTOPE MEASUREMENTS IN WATER, “Isotopes Environ. Health Stud.,” (India), 1998, Vol. 34, p. 349-369.